IPPCR 2019 Overview of Hypothesis Testing Part 1 of 5

>> Paul Wakim: Hello, I'm Paul Wakim. I'm chief of the Biostatistics and Clinical Epidemiology Service at the NIH Clinical Center, and this module is about hypothesis testing. We're going to cover a lot of concepts on hypothesis testing, and this module is -- corresponds to chapter 24 of the book "Principles and Practice of Clinical Research,"

the fourth edition. So, it's a five-part -- we're going to have five sections, five-part module, and we're going to talk about, in part one, statistical inference and confidence intervals. In part two, p-value and Bayesian approach. In part three we're going to talk about the American Statistical Association and the reporting of p-values. In part four, subgroup analysis and interaction.

And in part five, superiority versus non-inferiority versus equivalence and multiple comparisons. As you can see, I dedicated two parts to p-values. And that's because I think it's an extremely important concept, and I think it's most widely used and most widely misused statistical concept. And so, therefore, I thought it would be very important

to really clarify this concept of p-values. So, let's start with statistical inference and confidence interval, which is really this segment. So, what're we going to talk about -- what is statistical inference? What does that mean? It means you select, in the statistical concept -- in statistical context -- you select a representative sample

from the population of interest, you analyze the sample data, and you draw conclusion about the population based on results from the sample. So, I'm going to talk about the typical setting in statistical inference, the typical setting. And, here I talk about non-Bayesian. Most of this module on hypothesis testing is going to be the non-Bayesian approach.